[WIP] Support hpo in distributed #524
Open
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
rayrayraykk
commented
Feb 15, 2023
| if sender == self.server_id and self._cfg.hpo.use: | ||
| logger.info("Receive a new `cfg`, and start to reinitialize.") | ||
| self._cfg.defrost() | ||
| # TODO: Some var might remain unchanged |
Collaborator
Author
There was a problem hiding this comment.
We should add a reset() function to solve this.
rayrayraykk
commented
Feb 16, 2023
| config=init_cfg.clone(), | ||
| client_configs=client_cfgs) | ||
| _ = runner.run() | ||
| if init_cfg.hpo.use: |
Collaborator
Author
There was a problem hiding this comment.
We'd better use a fake optimizer to wrap this runner in the client to avoid a mismatch in sampled configuration.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
I add a function of sending
cfgbefore starting for run HPO in distributed mode. But it's not enough to changeself._cfgwhen the client receives a new configuration because some instances are created based on oldcfg, and we need to add a reset method for the worker to achieve this, just as thetrainerdoes, seeFederatedScope/federatedscope/core/trainers/trainer.py
Line 77 in b5870f7
And due to the limitations of gRPC, some vFL model and feature engineering methods can't work properly for the type of message might vary. In order to send such data, we may need a module for type conversion.
@xieyxclack @joneswong